Digital twin-driven power distribution network line loss analysis method and system

Abstract

The invention discloses a digital twinning driven power distribution network line loss analysis method and system, and relates to the technical field of power grids, and the method comprises the steps: carrying out the multi-dimensional line loss feature fitting of a power distribution network according to a distribution network control strategy based on a digital twinning assembly, building a multivariate line loss feature vector, carrying out the abnormality tracing, and building a line loss abnormality tracing atlas; adjusting the distribution network control strategy according to the line loss abnormity tracing atlas to obtain a first distribution network control adjustment domain; establishing a distribution network adjustment guide space, performing multi-level guide propagation on the first distribution network control adjustment domain, and establishing a second distribution network control adjustment domain; and performing multi-level optimization on the second distribution network control regulation domain, determining a distribution network control optimization result, and performing optimization management on the distribution network based on the distribution network control optimization result. The technical problems that in the prior art, power distribution network line loss is difficult to accurately model, abnormal tracing is insufficient, and global optimization regulation and control are lacked are solved, and the technical effects of reducing the line loss and improving the operation efficiency, safety and reliability of the power distribution network are achieved.

Description

Technical Field

[0001] This invention relates to the field of power grid technology, and more specifically to a method and system for analyzing line losses in distribution networks driven by digital twins. Background Technology

[0002] During operation, power distribution networks experience varying degrees of line loss due to factors such as conductor resistance, equipment performance degradation, and inadequate management practices. Common analytical methods rely on empirical models or calculations of single parameters, which fail to reflect the true characteristics of the distribution network under multi-dimensional and complex operating conditions, resulting in insufficient accuracy in line loss modeling. Line loss anomalies are often caused by the coupling of multiple factors, and the lack of a systematic tracing mechanism makes it difficult to pinpoint the root cause of problems in a timely manner, increasing the difficulty of operation and maintenance. Existing control methods focus on local optimization and lack comprehensive regulation capabilities across the entire network, failing to meet the demands of power distribution networks for efficient, safe, and intelligent management. Summary of the Invention

[0003] This application provides a digital twin-driven method and system for analyzing distribution network line losses, which addresses the technical problems in existing technologies such as difficulty in accurately modeling distribution network line losses, insufficient anomaly tracing, and lack of global optimization and control.

[0004] In view of the above problems, this application provides a digital twin-driven method and system for analyzing line losses in distribution networks.

[0005] The first aspect of this application provides a digital twin-driven method for analyzing line losses in distribution networks, the method comprising: Based on digital twin components, multi-dimensional line loss characteristics of the distribution network are fitted according to the distribution network control strategy to establish a multi-dimensional line loss feature vector; anomaly tracing is performed based on the multi-dimensional line loss feature vector to establish a line loss anomaly tracing map; the distribution network control strategy is adjusted according to the line loss anomaly tracing map to obtain a first distribution network control regulation domain; the first distribution network control regulation domain is guided and analyzed according to a multi-dimensional line loss risk prediction model to establish a distribution network regulation guidance space; the first distribution network control regulation domain is guided and propagated at multiple levels according to the distribution network regulation guidance space to establish a second distribution network control regulation domain; the second distribution network control regulation domain is optimized at multiple levels according to the multi-dimensional line loss risk prediction model to determine the distribution network control optimization result, and the distribution network is optimized and managed based on the distribution network control optimization result.

[0006] A second aspect of this application provides a digital twin-driven distribution network line loss analysis system, the system comprising: The feature fitting module 11 is used to perform multi-dimensional line loss feature fitting on the distribution network based on the digital twin component and the distribution network control strategy to establish a multi-dimensional line loss feature vector; the anomaly tracing module 12 is used to perform anomaly tracing based on the multi-dimensional line loss feature vector to establish a line loss anomaly tracing map; the adjustment module 13 is used to adjust the distribution network control strategy based on the line loss anomaly tracing map to obtain a first distribution network control adjustment domain; the guidance analysis module 14 is used to perform distribution network adjustment guidance analysis on the first distribution network control adjustment domain based on the multi-dimensional line loss risk prediction model to establish a distribution network adjustment guidance space; the guidance propagation module 15 is used to perform multi-level guidance propagation on the first distribution network control adjustment domain based on the distribution network adjustment guidance space to establish a second distribution network control adjustment domain; the optimization management module 16 is used to perform multi-level optimization on the second distribution network control adjustment domain based on the multi-dimensional line loss risk prediction model to determine the distribution network control optimization result, and perform optimization management on the distribution network based on the distribution network control optimization result.

[0007] One or more technical solutions provided in this application have at least the following technical effects or advantages: This application, based on a digital twin component, performs multi-dimensional line loss feature fitting on the distribution network according to the distribution network control strategy to establish a multi-element line loss feature vector; performs anomaly tracing based on the multi-element line loss feature vector to establish a line loss anomaly tracing map; adjusts the distribution network control strategy according to the line loss anomaly tracing map to obtain a first distribution network control regulation domain; performs distribution network regulation guidance analysis on the first distribution network control regulation domain according to a multi-dimensional line loss risk prediction model to establish a distribution network regulation guidance space; performs multi-level guidance propagation on the first distribution network control regulation domain according to the distribution network regulation guidance space to establish a second distribution network control regulation domain; performs multi-level optimization on the second distribution network control regulation domain according to the multi-dimensional line loss risk prediction model to determine the distribution network control optimization result, and optimizes the distribution network management based on the distribution network control optimization result. This invention addresses the technical problems of inaccurate modeling of distribution network line losses, insufficient anomaly tracing, and lack of global optimization and control in existing technologies. By constructing a virtual distribution network model, generating multi-dimensional line loss feature vectors, establishing a line loss anomaly tracing map, and combining it with a multi-dimensional line loss risk prediction model for multi-level adjustment and optimization, this invention achieves accurate identification, risk prediction, and intelligent optimization management of line losses, thereby effectively reducing line losses and improving the operational efficiency and safety reliability of the distribution network. Attached Figure Description

[0008] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0009] Figure 1 A schematic flowchart of the digital twin-driven distribution network line loss analysis method provided in this application embodiment; Figure 2 A schematic diagram of the structure of a digital twin-driven distribution network line loss analysis system provided in an embodiment of this application.

[0010] Figure labeling: Feature fitting module 11, Anomaly tracing module 12, Adjustment module 13, Guided analysis module 14, Guided reproduction module 15, Optimization management module 16. Detailed Implementation

[0011] This application provides a digital twin-driven method and system for analyzing distribution network line losses. It addresses the technical problems in existing technologies, such as difficulty in accurately modeling distribution network line losses, insufficient anomaly tracing, and a lack of global optimization and control. By constructing a virtual distribution network model, generating multi-dimensional line loss feature vectors, establishing a line loss anomaly tracing map, and combining it with a multi-dimensional line loss risk prediction model for multi-level adjustment and optimization, it achieves accurate identification, risk prediction, and intelligent optimization management of line losses. This results in effectively reducing line losses and improving the operational efficiency and reliability of the distribution network.

[0012] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0013] It should be noted that any variation of the terms "comprising" and "having" is intended to cover non-exclusive inclusion, for example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units that are explicitly listed, but may include other steps or modules that are not explicitly listed or that are inherent to such processes, methods, products, or devices.

[0014] Example 1, as Figure 1 As shown, this application provides a digital twin-driven method for analyzing line losses in distribution networks, the method comprising: Step S100: Based on the digital twin component, perform multi-dimensional line loss feature fitting on the distribution network according to the distribution network control strategy, and establish a multi-dimensional line loss feature vector.

[0015] In this embodiment, when performing multi-dimensional line loss characteristic fitting of the distribution network based on the distribution network control strategy using a digital twin component, a virtual distribution network model is first constructed, and a twin simulation is performed in this virtual distribution network model according to the distribution network control strategy to obtain the distribution network fitting dataset. The distribution network control strategy is a pre-prepared strategy to be executed.

[0016] Subsequently, multi-dimensional line loss feature identification was performed on the distribution network fitting dataset. Technical line loss feature vectors, management line loss feature vectors, fixed line loss feature vectors, and variable line loss feature vectors were extracted respectively. These feature vectors were then fused to form a multi-dimensional line loss feature vector.

[0017] Furthermore, the method provided in the application embodiments, based on a digital twin component, performs multi-dimensional line loss feature fitting on the distribution network according to the distribution network control strategy to establish a multi-dimensional line loss feature vector, and further includes: Based on the digital twin component, a virtual distribution network model of the distribution network is constructed; based on the digital twin component, a twin simulation of the virtual distribution network model is performed according to the distribution network control strategy to obtain a distribution network fitting dataset; technical line loss features are identified based on the distribution network fitting dataset to establish a technical line loss feature vector; management line loss features are identified based on the distribution network fitting dataset to establish a management line loss feature vector; fixed line loss features are identified based on the distribution network fitting dataset to establish a fixed line loss feature vector; variable line loss features are identified based on the distribution network fitting dataset to establish a variable line loss feature vector; and the multivariate line loss feature vector is generated by combining the technical line loss feature vector, the management line loss feature vector, and the fixed line loss feature vector.

[0018] In this embodiment, the digital twin component is a virtualization modeling and simulation tool. By collecting structural information and operational data of the power grid, it constructs a model in digital space that corresponds to the real distribution network and can simulate the operating state of the power grid in real time. When constructing a virtual distribution network model based on the digital twin component, the information on the lines, nodes, and equipment of the distribution network is first organized, and then the operating parameters are input into the virtual environment, thereby generating a virtual distribution network model that is consistent with the real power grid in structure and state.

[0019] Subsequently, based on digital twin components, when performing twin simulations on the virtual distribution network model according to the distribution network control strategy, the distribution network control strategy is input into the virtual distribution network model. The distribution network control strategy refers to the scheduling and management strategies to be executed during power grid operation, such as voltage and reactive power optimization strategies, power flow allocation strategies, or load switching strategies. The twin simulation generates a distribution network fitting dataset covering different operating conditions and time sequences by executing these strategies. This distribution network fitting dataset includes operating indicators such as current, voltage, and power loss.

[0020] Next, technical line loss features are identified based on the distribution network fitting dataset. Technical line loss is energy loss caused by the inherent physical characteristics of electrical components, including line resistance heating, transformer copper loss, and iron loss. By performing power flow calculations and loss decomposition on the distribution network fitting dataset, parameters such as current square, line impedance, and power loss are extracted to establish a technical line loss feature vector.

[0021] Subsequently, the characteristics of managed line losses are identified. Managed line losses are mainly caused by factors such as metering deviations, incomplete data collection, and electricity theft. By comparing the fitted dataset of the distribution network with the metering statistics, and combining anomaly detection methods to identify data differences, characteristic indicators such as metering error rate and load pattern deviation are extracted to form a feature vector of managed line losses.

[0022] Next, fixed line loss features are extracted. Fixed line loss refers to long-term losses that do not change significantly with operating conditions, such as equipment no-load losses and station power consumption. By performing baseline estimation and statistical analysis on fitted data under low load or no-load conditions, the stable losses are identified, and a fixed line loss feature vector is generated.

[0023] Finally, variable line loss features are extracted. Variable line loss fluctuates with changes in load level and operating conditions, such as increased line resistance and losses due to increased load. By analyzing the loss changes under different load curves in the distribution network fitting dataset, and combining differential analysis and sensitivity calculations, a variable line loss feature vector is established.

[0024] After identifying the above four types of feature vectors, the technical line loss feature vector, management line loss feature vector, fixed line loss feature vector, and variable line loss feature vector are fused to generate a multi-dimensional line loss feature vector.

[0025] Step S200: Perform anomaly tracing based on the multivariate line loss feature vector to establish a line loss anomaly tracing map.

[0026] In this embodiment, when tracing anomalies based on multi-dimensional line loss feature vectors, anomaly identification is first performed on the multi-dimensional line loss feature vectors to distinguish technical line loss anomalies, management line loss anomalies, fixed line loss anomalies, and variable line loss anomalies. Then, for technical anomalies, a technical line loss anomaly tracing model is established using fault tree learning, and causal analysis is performed in conjunction with the technical line loss feature vectors to obtain the technical line loss anomaly tracing results. For management, fixed, and variable anomalies, fault tree tracing analysis is also used for causal tracing, yielding management line loss anomaly tracing results, fixed line loss anomaly tracing results, and variable line loss anomaly tracing results, respectively. After summarizing the above results, the different types of tracing results are fused and expressed to construct a line loss anomaly tracing map.

[0027] Furthermore, in the method provided in the application embodiments, the abnormal tracing based on the multi-dimensional line loss feature vector to establish a line loss abnormal tracing map further includes: Anomaly identification is performed based on the multivariate line loss feature vector to determine technical line loss anomaly features, management line loss anomaly features, fixed line loss anomaly features, and variable line loss anomaly features. Fault tree learning is performed based on the technical line loss anomaly event set to establish a technical line loss anomaly tracing model. Based on the technical line loss feature vector, the technical line loss anomaly features are traced and analyzed according to the technical line loss anomaly tracing model to obtain technical line loss anomaly tracing results. Fault tree tracing analysis is performed on the management line loss anomaly features, the fixed line loss anomaly features, and the variable line loss anomaly features respectively to obtain management line loss anomaly tracing results, fixed line loss anomaly tracing results, and variable line loss anomaly tracing results. Based on the technical line loss anomaly tracing results, the management line loss anomaly tracing results, the fixed line loss anomaly tracing results, and the variable line loss anomaly tracing results, the line loss anomaly tracing map is generated.

[0028] In this embodiment, when identifying anomalies based on a multi-dimensional line loss feature vector, a statistical threshold determination method is used to compare the values ​​of each dimension of the feature vector in the multi-dimensional line loss feature vector with a preset threshold, and the portion exceeding the threshold range is identified as an anomaly. This method distinguishes between technical line loss anomaly features, management line loss anomaly features, fixed line loss anomaly features, and variable line loss anomaly features.

[0029] Subsequently, fault tree learning was performed based on the set of abnormal technical line loss events. This set of events, derived from historical operational data, included anomalies such as line overload, electrical faults, and equipment aging. By employing the fault tree learning method—that is, by decomposing the logical relationships between events and causes layer by layer to establish a causal model—a fault tree model was ultimately obtained to trace abnormal technical line loss events.

[0030] Next, based on the existing technical line loss feature vector, the currently identified technical line loss anomalies are input into the technical line loss anomaly tracing model. A causal path backtracking method is used to trace the anomalies layer by layer, locating possible root cause events, such as increased conductor resistance or abnormal transformer losses, thus obtaining the technical line loss anomaly tracing results. When analyzing the characteristics of management line loss anomalies, a fault tree tracing analysis method is used, with metering equipment errors, data acquisition system failures, and abnormal electricity consumption behavior as basic event nodes. Through layer-by-layer deduction, the root causes leading to management line loss anomalies are determined, forming the management line loss anomaly tracing results. When analyzing the characteristics of fixed line loss anomalies, the same fault tree tracing method is used, with long-term no-load loss anomalies and high station power consumption as event inputs, identifying their causal chains, and obtaining the fixed line loss anomaly tracing results. When analyzing the characteristics of variable line loss anomalies, combined with load fluctuation characteristics, fault tree tracing analysis is used, with sudden load increases, abnormal environmental conditions, and unreasonable control strategies as bottom-level event nodes, gradually tracing the source of the anomalies, thus obtaining the variable line loss anomaly tracing results.

[0031] Finally, the results of technical line loss anomaly tracing, management line loss anomaly tracing, fixed line loss anomaly tracing, and variable line loss anomaly tracing are integrated. Using a graph representation method, different anomaly characteristics, anomaly events, and causal relationships are displayed in the form of nodes and edges to generate a line loss anomaly tracing graph.

[0032] Step S300: Adjust the distribution network control strategy according to the line loss anomaly tracing map to obtain the first distribution network control adjustment domain.

[0033] In this embodiment of the application, when adjusting the distribution network control strategy based on the line loss anomaly tracing map, the technical line loss anomaly characteristics, management line loss anomaly characteristics, fixed line loss anomaly characteristics, and variable line loss anomaly characteristics in the line loss anomaly tracing map are first used to identify the parameter links related to the anomaly in the current distribution network control strategy through the difference matching method, such as voltage setpoint, reactive power distribution method, load switching arrangement, and power flow distribution scheme, thereby determining the specific control object that needs to be adjusted.

[0034] After identifying the problem link, according to the causal relationship provided by the line loss anomaly traceability map, the existing distribution network control strategy is corrected using parameter correction rules. For example, when the technical line loss anomaly traceability result points to line overload, the corresponding distribution network control strategy adjusts the power flow distribution ratio to reduce the line current; when the management line loss anomaly traceability result points to measurement error, the distribution network control strategy adds verification measures in the dispatching link; when the fixed line loss anomaly traceability result shows that the no-load loss of the equipment is too high, the equipment switching constraints are increased; when the variable line loss anomaly traceability result indicates that the load fluctuation is too large, the load transfer and reactive power compensation parameters are adjusted. Through these methods, correction rules matching the anomaly traceability results are formed.

[0035] After the application of the correction rules is completed, different correction parameters are arranged and combined using the combination generation method to obtain multiple executable distribution network adjustment schemes. Finally, these multiple distribution network adjustment schemes are assembled to form the first distribution network control adjustment domain.

[0036] Step S400: Perform distribution network adjustment guidance analysis on the first distribution network control adjustment domain according to the multi-dimensional line loss risk prediction model, and establish a distribution network adjustment guidance space.

[0037] In the embodiment of the present application, when performing distribution network adjustment guidance analysis on the first distribution network control adjustment domain according to the pre-trained multi-dimensional line loss risk prediction model, first, the multi-dimensional line loss risk prediction model performs risk iterative optimization on the first distribution network control adjustment domain dimension by dimension, and sequentially obtains the first distribution network adjustment guidance vector corresponding to the technical line loss risk iterative optimization result, the second distribution network adjustment guidance vector corresponding to the management line loss risk iterative optimization result, the third distribution network adjustment guidance vector corresponding to the fixed line loss risk iterative optimization result, and the fourth distribution network adjustment guidance vector corresponding to the variable line loss risk iterative optimization result, and unify and summarize the above four types of vectors to establish a distribution network adjustment guidance space.

[0038] Furthermore, the method provided in the embodiment of the application further includes: The multi-dimensional line loss risk prediction model includes multi-dimensional line loss risk prediction indicators, and the multi-dimensional line loss risk prediction indicators include technical line loss risk, management line loss risk, fixed line loss risk, and variable line loss risk.

[0039] In the embodiment of the present application, the multi-dimensional line loss risk prediction model includes multi-dimensional line loss risk prediction indicators, and the multi-dimensional line loss risk prediction indicators include technical line loss risk, management line loss risk, fixed line loss risk, and variable line loss risk.

[0040] In the training process of the multidimensional line loss risk prediction model, pre-prepared historical data is used. This historical data is divided into input data and output data. The input data consists of distribution network control and regulation schemes under different scenarios; the output data consists of risk coefficients pre-labeled by technical experts for each distribution network control and regulation scheme, including technical line loss risk coefficients, management line loss risk coefficients, fixed line loss risk coefficients, and variable line loss risk coefficients. By mapping the input data to the output data, machine learning methods are used to establish a mapping relationship between input features and risk coefficients. Parameters are optimized through iterative training, ultimately completing the training of the multidimensional line loss risk prediction model.

[0041] Furthermore, in the method provided in the application embodiments, the method of performing distribution network regulation guidance analysis on the first distribution network control and regulation domain according to the multidimensional line loss risk prediction model to establish a distribution network regulation guidance space also includes: Based on the multidimensional line loss risk prediction model, the technical line loss risk of the first distribution network control and regulation domain is iteratively optimized to obtain a first distribution network regulation guidance vector; based on the multidimensional line loss risk prediction model, the management line loss risk of the first distribution network control and regulation domain is iteratively optimized to obtain a second distribution network regulation guidance vector; based on the multidimensional line loss risk prediction model, the fixed line loss risk of the first distribution network control and regulation domain is iteratively optimized to obtain a third distribution network regulation guidance vector; based on the multidimensional line loss risk prediction model, the variable line loss risk of the first distribution network control and regulation domain is iteratively optimized to obtain a fourth distribution network regulation guidance vector; the first distribution network regulation guidance vector, the second distribution network regulation guidance vector, the third distribution network regulation guidance vector, and the fourth distribution network regulation guidance vector are added to the distribution network regulation guidance space.

[0042] In this embodiment, firstly, for the technical line loss risk, each distribution network control and regulation scheme in the first distribution network control and regulation domain is sequentially input into the multi-dimensional line loss risk prediction model to obtain the corresponding technical line loss risk coefficient. By comparing the magnitudes of the technical line loss risk coefficients of each distribution network control and regulation scheme, the distribution network control and regulation scheme corresponding to the smallest technical line loss risk coefficient is selected, and this distribution network control and regulation scheme is used as the first distribution network regulation guidance vector.

[0043] Subsequently, for the management line loss risk, the same method was used to predict all schemes in the first distribution network control and regulation domain, obtaining their respective management line loss risk coefficients. Through iterative comparison, the distribution network control and regulation scheme corresponding to the minimum management line loss risk coefficient was selected as the second distribution network regulation guidance vector, thereby clarifying the optimal regulation path under the management line loss risk dimension.

[0044] Next, the fixed line loss risk is addressed. Using all distribution network control and regulation schemes as input, the fixed line loss risk coefficients output by the multi-dimensional line loss risk prediction model are compared. The distribution network control and regulation scheme with the smallest coefficient is selected as the third distribution network regulation guidance vector to reflect the optimal selection under the fixed line loss dimension.

[0045] Finally, the variable line loss risk is analyzed. The variable line loss risk coefficient of each regulation scheme is predicted by a multi-dimensional line loss risk prediction model. The scheme with the smallest coefficient value is selected as the fourth distribution network regulation guidance vector, which is used to characterize the optimal scheme under load fluctuation and operating status changes.

[0046] After obtaining the four types of guiding vectors, the first distribution network regulation guiding vector, the second distribution network regulation guiding vector, the third distribution network regulation guiding vector, and the fourth distribution network regulation guiding vector are summarized and added to the distribution network regulation guiding space.

[0047] Step S500: Based on the distribution network regulation guidance space, perform multi-level guidance and propagation on the first distribution network control regulation domain to establish a second distribution network control regulation domain.

[0048] In this embodiment, when performing multi-level guided propagation of the first distribution network control and regulation domain based on the distribution network regulation guidance space, the first distribution network regulation guidance vector is first used for difference analysis to generate difference features, and through multiple mutations, a propagation guidance particle swarm is formed, thereby guiding the first distribution network regulation propagation domain. Subsequently, the second, third, and fourth distribution network regulation guidance vectors are used sequentially to guide propagation of the first distribution network control and regulation domain, respectively, generating the second, third, and fourth distribution network regulation propagation domains. After obtaining these four types of propagation domains, they are integrated and expanded with the first distribution network control and regulation domain to finally generate the second distribution network control and regulation domain.

[0049] Furthermore, in the method provided in the application embodiments, the method of multi-level guiding and propagating the first distribution network control and regulation domain according to the distribution network regulation and guidance space to establish a second distribution network control and regulation domain further includes: The first distribution network control and regulation domain is analyzed for differences based on the first distribution network regulation guidance vector to obtain a first difference feature particle group. Multiple mutations are performed on the first difference feature particle group to obtain a first breeding guidance particle group. This first breeding guidance particle group guides the first distribution network control and regulation domain to reproduce, resulting in a first distribution network regulation breeding domain. The first distribution network control and regulation domain is then bred using a second distribution network regulation guidance vector to obtain a second distribution network regulation breeding domain. The first distribution network control and regulation domain is then bred using a third distribution network regulation guidance vector to obtain a third distribution network regulation breeding domain. The first distribution network control and regulation domain is then bred using a fourth distribution network regulation guidance vector to obtain a fourth distribution network regulation breeding domain. Finally, the first distribution network control and regulation domain is expanded using the first, second, third, and fourth distribution network regulation breeding domains to generate the second distribution network control and regulation domain.

[0050] In this embodiment, the first distribution network control and regulation domain is first analyzed for differences based on the first distribution network regulation guidance vector. The parameter differences between each distribution network control and regulation scheme and the first distribution network regulation guidance vector are calculated using the absolute difference measurement method. A preset threshold is introduced for screening, and schemes with differences greater than the preset threshold and satisfying preset operating constraints are retained. These schemes are represented in particle form and constitute the first difference feature particle group.

[0051] Next, based on the first differential characteristic particle swarm, a random perturbation method is used to apply small changes to key control parameters within the allowable range. After the perturbation, out-of-bounds truncation and repeated elimination are performed to generate a first breeding guide particle swarm with directionality and diversity. Then, using this first breeding guide particle swarm as the starting point, directional incremental updates are performed according to the first distribution network regulation guide vector to generate new regulation schemes, which are then judged based on loss calculation values. Specifically, the loss calculation value is obtained by multiplying the line resistance by the square of the current to obtain the line loss, and then combining this with the equivalent formulas for transformer copper loss and iron loss to obtain the equipment loss. The sum of these two values ​​is used as the loss calculation value for each scheme. Schemes that are below the preset threshold and meet the operating constraints such as voltage and current are retained, forming the first distribution network regulation breeding domain.

[0052] Then, the first distribution network control and regulation domain is guided and propagated based on the second distribution network regulation guidance vector. A new set of schemes is generated between the original scheme and the second distribution network regulation guidance vector using a linear interpolation method. The loss calculation value of each scheme is calculated, and the schemes that meet the conditions are selected and retained to form the second distribution network regulation propagation domain.

[0053] Next, the control and regulation domain of the first distribution network is guided and propagated based on the third distribution network regulation guidance vector. The parameters related to fixed line loss are adjusted using a step-size fine-tuning method, and the loss calculation values ​​before and after the adjustment are calculated respectively. The scheme with significant improvement is retained by judging through a preset threshold, and the third distribution network regulation propagation domain is obtained.

[0054] Subsequently, the first distribution network control and regulation domain is guided and propagated based on the fourth distribution network regulation guidance vector. A multi-scenario sampling method is used to generate candidate schemes under different load and environmental conditions. A simplified time-series simulation method is used to calculate the loss calculation values ​​of each scheme under different scenarios. A preset threshold is used as a screening condition to obtain the fourth distribution network regulation propagation domain.

[0055] Finally, the first distribution network regulation breeding domain, the second distribution network regulation breeding domain, the third distribution network regulation breeding domain, and the fourth distribution network regulation breeding domain are merged with the first distribution network control and regulation domain. The duplicate and infeasible schemes are eliminated by deduplication and threshold filtering methods to generate the second distribution network control and regulation domain.

[0056] Step S600: Perform multi-level optimization on the second distribution network control and regulation domain according to the multi-dimensional line loss risk prediction model, determine the distribution network control optimization result, and optimize the distribution network management based on the distribution network control optimization result.

[0057] In this embodiment, when performing multi-level optimization of the second distribution network control and regulation domain based on the multi-dimensional line loss risk prediction model, the multi-dimensional line loss risk prediction model is first used to optimize the line loss risk constraint of the second distribution network control and regulation domain to obtain the third distribution network control and regulation domain. Then, a global line loss risk analysis model is constructed based on the multi-dimensional line loss risk weights, and global line loss risk analysis is performed in the third distribution network control and regulation domain to obtain the corresponding global line loss risk sequence. Next, a global line loss risk threshold is introduced to perform optimization identification of the third distribution network control and regulation domain, filtering to form the fourth distribution network control and regulation domain. Finally, energy consumption minimization optimization is performed based on the fourth distribution network control and regulation domain to determine the final distribution network control optimization result.

[0058] After obtaining the distribution network control optimization results, these results are applied to the actual operation of the distribution network. By adjusting voltage setpoints, optimizing reactive power allocation, implementing power flow path reconfiguration, and improving load transfer strategies, the optimized management of the distribution network is achieved.

[0059] Furthermore, in the method provided in the application embodiments, the method of performing multi-level optimization on the second distribution network control and regulation domain according to the multi-dimensional line loss risk prediction model to determine the distribution network control optimization result further includes: The second distribution network control and regulation domain is optimized based on the multidimensional line loss risk prediction model to obtain the third distribution network control and regulation domain. A global line loss risk analysis model is constructed based on the multidimensional line loss risk weights. Global line loss risk analysis is performed on the third distribution network control and regulation domain based on the global line loss risk analysis model to obtain a global line loss risk sequence. Based on the global line loss risk sequence and a global line loss risk threshold, the third distribution network control and regulation domain is optimized to obtain the fourth distribution network control and regulation domain. Energy consumption minimization optimization is performed on the fourth distribution network control and regulation domain to generate the distribution network control optimization result.

[0060] Furthermore, the method provided in the application embodiments also includes: The multidimensional line loss risk weights include technical line loss risk weights, management line loss risk weights, fixed line loss risk weights, and variable line loss risk weights.

[0061] In this embodiment, when optimizing the line loss risk constraints of the second distribution network control and regulation domain based on the multidimensional line loss risk prediction model, the distribution network control and regulation schemes in the second distribution network control and regulation domain are first extracted one by one, and the corresponding line loss risk prediction matrix is ​​calculated based on the multidimensional line loss risk prediction model. Subsequently, a line loss risk constraint matrix is ​​constructed based on preset technical line loss risk constraints, management line loss risk constraints, fixed line loss risk constraints, and variable line loss risk constraints to determine the risk level of each candidate scheme. If a certain line loss risk prediction matrix satisfies the conditions of the line loss risk constraint matrix, the corresponding distribution network control and regulation scheme is included in the third distribution network control and regulation domain. By sequentially performing the above screening and updating process on all schemes in the second distribution network control and regulation domain, a third distribution network control and regulation domain that satisfies the risk constraint conditions is gradually formed.

[0062] Next, a global line loss risk analysis model is constructed based on the multi-dimensional line loss risk weights. These multi-dimensional line loss risk weights include technical line loss risk weights, management line loss risk weights, fixed line loss risk weights, and variable line loss risk weights. The resulting global line loss risk analysis model is a weighted calculation model, expressed by the formula: Global Line Loss Risk Coefficient = Technical Line Loss Risk Coefficient × Technical Line Loss Risk Weight + Management Line Loss Risk Coefficient × Management Line Loss Risk Weight + Fixed Line Loss Risk Coefficient × Fixed Line Loss Risk Weight + Variable Line Loss Risk Coefficient × Variable Line Loss Risk Weight.

[0063] Subsequently, a global line loss risk analysis was performed on the third distribution network control and regulation domain based on the global line loss risk analysis model. In this process, for each distribution network control and regulation scheme in the third distribution network control and regulation domain, the corresponding global line loss risk coefficient was calculated using the global line loss risk analysis model. The global line loss risk coefficients of all candidate schemes were then grouped into a sequence to form a global line loss risk sequence.

[0064] After obtaining the global line loss risk sequence, the third distribution network control and regulation domain is optimized based on the set global line loss risk threshold. The threshold discrimination method is used to check the global line loss risk coefficient in the global line loss risk sequence one by one. If it is less than the global line loss risk threshold, the corresponding distribution network control and regulation scheme is retained in the fourth distribution network control and regulation domain.

[0065] Finally, energy consumption minimization optimization is performed based on the fourth distribution network control and regulation domain. This process begins by calculating the energy consumption of each distribution network control and regulation scheme within this domain. This calculation includes line energy consumption calculated by multiplying the line resistance by the square of the current, equipment energy consumption calculated using formulas for transformer copper and iron losses, and time-series energy consumption derived from load demands during different operating periods. These three factors are then combined to form the energy consumption index for each scheme. Subsequently, using the energy consumption index as the objective function, an optimization method is employed to screen candidate schemes within the fourth distribution network control and regulation domain. Schemes that do not meet voltage, current, and power supply safety constraints are eliminated, and the distribution network control and regulation scheme with the minimum energy consumption index is determined, generating the final distribution network control optimization result.

[0066] Furthermore, in the method provided in the application embodiments, the method of optimizing the second distribution network control and regulation domain based on the multidimensional line loss risk prediction model to obtain the third distribution network control and regulation domain further includes: The b-th distribution network control and regulation scheme is extracted from the second distribution network control and regulation domain, where b is a positive integer. Based on the b-th distribution network control and regulation scheme, the b-th line loss risk prediction matrix is ​​obtained according to the multi-dimensional line loss risk prediction model. A line loss risk constraint matrix is ​​constructed based on technical line loss risk constraints, management line loss risk constraints, fixed line loss risk constraints, and variable line loss risk constraints. If the b-th line loss risk prediction matrix satisfies the line loss risk constraint matrix, the b-th distribution network control and regulation scheme is added to the third distribution network control and regulation domain. Based on the line loss risk constraint matrix, the second distribution network control and regulation domain is further optimized according to the multi-dimensional line loss risk prediction model, and the third distribution network control and regulation domain is updated.

[0067] In this embodiment, when optimizing the line loss risk constraint for the second distribution network control and regulation domain, the b-th distribution network control and regulation scheme is first extracted from the second distribution network control and regulation domain, where b is a positive integer. Based on this b-th distribution network control and regulation scheme, it is input into a multi-dimensional line loss risk prediction model for calculation, resulting in the corresponding b-th line loss risk prediction matrix. This line loss risk prediction matrix consists of four elements: the technical line loss risk coefficient, the management line loss risk coefficient, the fixed line loss risk coefficient, and the variable line loss risk coefficient corresponding to the b-th distribution network control and regulation scheme.

[0068] Subsequently, based on preset technical line loss risk constraints, management line loss risk constraints, fixed line loss risk constraints, and variable line loss risk constraints, a line loss risk constraint matrix is ​​constructed. This matrix defines the allowable threshold limits under each risk dimension. Using a matrix comparison method, the b-th line loss risk prediction matrix is ​​compared element-by-element with the line loss risk constraint matrix. If all risk coefficients in the b-th line loss risk prediction matrix satisfy the corresponding constraints, the b-th distribution network control and regulation scheme is determined to meet the risk constraints, and the scheme is added to the third distribution network control and regulation domain.

[0069] After completing the above operations, based on the line loss risk constraint matrix, the same calculations and judgments are performed on other distribution network control and regulation schemes in the second distribution network control and regulation domain using the multi-dimensional line loss risk prediction model, updating the third distribution network control and regulation domain one by one. Ultimately, the third distribution network control and regulation domain only includes those distribution network control and regulation schemes calculated by the multi-dimensional line loss risk prediction model and meeting the requirements of the line loss risk constraint matrix.

[0070] In summary, the embodiments of this application have at least the following technical effects: This application, based on a digital twin component, performs multi-dimensional line loss feature fitting on the distribution network according to the distribution network control strategy to establish a multi-element line loss feature vector; performs anomaly tracing based on the multi-element line loss feature vector to establish a line loss anomaly tracing map; adjusts the distribution network control strategy according to the line loss anomaly tracing map to obtain a first distribution network control regulation domain; performs distribution network regulation guidance analysis on the first distribution network control regulation domain according to a multi-dimensional line loss risk prediction model to establish a distribution network regulation guidance space; performs multi-level guidance propagation on the first distribution network control regulation domain according to the distribution network regulation guidance space to establish a second distribution network control regulation domain; performs multi-level optimization on the second distribution network control regulation domain according to the multi-dimensional line loss risk prediction model to determine the distribution network control optimization result, and optimizes the distribution network management based on the distribution network control optimization result. This invention addresses the technical problems of inaccurate modeling of distribution network line losses, insufficient anomaly tracing, and lack of global optimization and control in existing technologies. By constructing a virtual distribution network model, generating multi-dimensional line loss feature vectors, establishing a line loss anomaly tracing map, and combining it with a multi-dimensional line loss risk prediction model for multi-level adjustment and optimization, this invention achieves accurate identification, risk prediction, and intelligent optimization management of line losses, thereby effectively reducing line losses and improving the operational efficiency and safety reliability of the distribution network.

[0071] Example 2, based on the same inventive concept as the digital twin-driven distribution network line loss analysis method in the foregoing examples, such as... Figure 2 As shown, this application provides a digital twin-driven distribution network line loss analysis system. The system and method embodiments in this application are based on the same inventive concept. The system includes: The feature fitting module 11 is used to perform multi-dimensional line loss feature fitting on the distribution network based on the digital twin component and the distribution network control strategy to establish a multi-dimensional line loss feature vector; the anomaly tracing module 12 is used to perform anomaly tracing based on the multi-dimensional line loss feature vector to establish a line loss anomaly tracing map; the adjustment module 13 is used to adjust the distribution network control strategy based on the line loss anomaly tracing map to obtain a first distribution network control adjustment domain; the guidance analysis module 14 is used to perform distribution network adjustment guidance analysis on the first distribution network control adjustment domain based on the multi-dimensional line loss risk prediction model to establish a distribution network adjustment guidance space; the guidance propagation module 15 is used to perform multi-level guidance propagation on the first distribution network control adjustment domain based on the distribution network adjustment guidance space to establish a second distribution network control adjustment domain; the optimization management module 16 is used to perform multi-level optimization on the second distribution network control adjustment domain based on the multi-dimensional line loss risk prediction model to determine the distribution network control optimization result, and perform optimization management on the distribution network based on the distribution network control optimization result.

[0072] Furthermore, the system is also used to implement the following functions: Based on the digital twin component, a virtual distribution network model of the distribution network is constructed; based on the digital twin component, a twin simulation of the virtual distribution network model is performed according to the distribution network control strategy to obtain a distribution network fitting dataset; technical line loss features are identified based on the distribution network fitting dataset to establish a technical line loss feature vector; management line loss features are identified based on the distribution network fitting dataset to establish a management line loss feature vector; fixed line loss features are identified based on the distribution network fitting dataset to establish a fixed line loss feature vector; variable line loss features are identified based on the distribution network fitting dataset to establish a variable line loss feature vector; and the multivariate line loss feature vector is generated by combining the technical line loss feature vector, the management line loss feature vector, and the fixed line loss feature vector.

[0073] Furthermore, the system is also used to implement the following functions: Anomaly identification is performed based on the multivariate line loss feature vector to determine technical line loss anomaly features, management line loss anomaly features, fixed line loss anomaly features, and variable line loss anomaly features. Fault tree learning is performed based on the technical line loss anomaly event set to establish a technical line loss anomaly tracing model. Based on the technical line loss feature vector, the technical line loss anomaly features are traced and analyzed according to the technical line loss anomaly tracing model to obtain technical line loss anomaly tracing results. Fault tree tracing analysis is performed on the management line loss anomaly features, the fixed line loss anomaly features, and the variable line loss anomaly features respectively to obtain management line loss anomaly tracing results, fixed line loss anomaly tracing results, and variable line loss anomaly tracing results. Based on the technical line loss anomaly tracing results, the management line loss anomaly tracing results, the fixed line loss anomaly tracing results, and the variable line loss anomaly tracing results, the line loss anomaly tracing map is generated.

[0074] Furthermore, the system is also used to implement the following functions: Based on the multidimensional line loss risk prediction model, the technical line loss risk of the first distribution network control and regulation domain is iteratively optimized to obtain a first distribution network regulation guidance vector; based on the multidimensional line loss risk prediction model, the management line loss risk of the first distribution network control and regulation domain is iteratively optimized to obtain a second distribution network regulation guidance vector; based on the multidimensional line loss risk prediction model, the fixed line loss risk of the first distribution network control and regulation domain is iteratively optimized to obtain a third distribution network regulation guidance vector; based on the multidimensional line loss risk prediction model, the variable line loss risk of the first distribution network control and regulation domain is iteratively optimized to obtain a fourth distribution network regulation guidance vector; the first distribution network regulation guidance vector, the second distribution network regulation guidance vector, the third distribution network regulation guidance vector, and the fourth distribution network regulation guidance vector are added to the distribution network regulation guidance space.

[0075] Furthermore, the system is also used to implement the following functions: The first distribution network control and regulation domain is analyzed for differences based on the first distribution network regulation guidance vector to obtain a first difference feature particle group. Multiple mutations are performed on the first difference feature particle group to obtain a first breeding guidance particle group. This first breeding guidance particle group guides the first distribution network control and regulation domain to reproduce, resulting in a first distribution network regulation breeding domain. The first distribution network control and regulation domain is then bred using a second distribution network regulation guidance vector to obtain a second distribution network regulation breeding domain. The first distribution network control and regulation domain is then bred using a third distribution network regulation guidance vector to obtain a third distribution network regulation breeding domain. The first distribution network control and regulation domain is then bred using a fourth distribution network regulation guidance vector to obtain a fourth distribution network regulation breeding domain. Finally, the first distribution network control and regulation domain is expanded using the first, second, third, and fourth distribution network regulation breeding domains to generate the second distribution network control and regulation domain.

[0076] Furthermore, the system is also used to implement the following functions: The second distribution network control and regulation domain is optimized based on the multidimensional line loss risk prediction model to obtain the third distribution network control and regulation domain. A global line loss risk analysis model is constructed based on the multidimensional line loss risk weights. Global line loss risk analysis is performed on the third distribution network control and regulation domain based on the global line loss risk analysis model to obtain a global line loss risk sequence. Based on the global line loss risk sequence and a global line loss risk threshold, the third distribution network control and regulation domain is optimized to obtain the fourth distribution network control and regulation domain. Energy consumption minimization optimization is performed on the fourth distribution network control and regulation domain to generate the distribution network control optimization result.

[0077] Furthermore, the system is also used to implement the following functions: The b-th distribution network control and regulation scheme is extracted from the second distribution network control and regulation domain, where b is a positive integer. Based on the b-th distribution network control and regulation scheme, the b-th line loss risk prediction matrix is ​​obtained according to the multi-dimensional line loss risk prediction model. A line loss risk constraint matrix is ​​constructed based on technical line loss risk constraints, management line loss risk constraints, fixed line loss risk constraints, and variable line loss risk constraints. If the b-th line loss risk prediction matrix satisfies the line loss risk constraint matrix, the b-th distribution network control and regulation scheme is added to the third distribution network control and regulation domain. Based on the line loss risk constraint matrix, the second distribution network control and regulation domain is further optimized according to the multi-dimensional line loss risk prediction model, and the third distribution network control and regulation domain is updated.

[0078] Furthermore, the system is also used to implement the following functions: The multidimensional line loss risk prediction model includes multidimensional line loss risk prediction indicators, which include technical line loss risk, management line loss risk, fixed line loss risk, and variable line loss risk.

[0079] Furthermore, the system is also used to implement the following functions: The multidimensional line loss risk weights include technical line loss risk weights, management line loss risk weights, fixed line loss risk weights, and variable line loss risk weights.

[0080] It should be noted that the order of the embodiments described above is for descriptive purposes only and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous.

[0081] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

[0082] This specification and accompanying drawings are merely illustrative examples of this application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Therefore, if such modifications and modifications fall within the scope of this application and its equivalents, this application intends to include such modifications and modifications.

Claims

1. A digital twin driven power distribution network line loss analysis method, characterized in that, The method comprises: Based on the digital twin component, the multi-dimensional line loss characteristics of the distribution network are fitted according to the distribution network control strategy, and a multi-element line loss characteristic vector is established; According to the multi-element line loss characteristic vector, abnormal backtracking is carried out, and a line loss abnormal backtracking map is established; According to the line loss abnormal backtracking map, the distribution network control strategy is adjusted to obtain a first distribution network control adjustment domain; According to the multi-dimensional line loss risk prediction model, the distribution network adjustment guide space is established by analyzing the first distribution network control adjustment domain; According to the multi-dimensional line loss risk prediction model, the second distribution network control adjustment domain is established by multi-level guided breeding of the first distribution network control adjustment domain; According to the multi-dimensional line loss risk prediction model, the second distribution network control adjustment domain is determined by multi-level optimization, and the distribution network is optimized and managed based on the distribution network control optimization result.

2. The digital twin driven power distribution network line loss analysis method of claim 1, wherein, Based on the digital twin component, the multi-dimensional line loss characteristics of the distribution network are fitted according to the distribution network control strategy, and a multi-element line loss characteristic vector is established, including: According to the digital twin component, a virtual distribution network model of the distribution network is constructed; Based on the digital twin component, the virtual distribution network model is simulated according to the distribution network control strategy to obtain a distribution network fitting data set; According to the distribution network fitting data set, technical line loss characteristic identification is carried out to establish a technical line loss characteristic vector; According to the distribution network fitting data set, management line loss characteristic identification is carried out to establish a management line loss characteristic vector; According to the distribution network fitting data set, fixed line loss characteristic identification is carried out to establish a fixed line loss characteristic vector; According to the distribution network fitting data set, variable line loss characteristic identification is carried out to establish a variable line loss characteristic vector, and the multi-element line loss characteristic vector is generated by combining the technical line loss characteristic vector, the management line loss characteristic vector and the fixed line loss characteristic vector.

3. The digital twin driven power distribution network line loss analysis method of claim 1, wherein, According to the multi-element line loss characteristic vector, abnormal backtracking is carried out, and a line loss abnormal backtracking map is established, including: According to the multi-element line loss characteristic vector, abnormal identification is carried out to determine technical line loss abnormal characteristics, management line loss abnormal characteristics, fixed line loss abnormal characteristics and variable line loss abnormal characteristics; According to the technical line loss abnormal event set, an accident tree is learned to establish a technical line loss abnormal backtracking model; Based on the technical line loss characteristic vector, the technical line loss abnormal characteristics are analyzed according to the technical line loss abnormal backtracking model to obtain a technical line loss abnormal backtracking result; The management line loss abnormal characteristics, the fixed line loss abnormal characteristics and the variable line loss abnormal characteristics are respectively subjected to accident tree backtracking analysis to obtain management line loss abnormal backtracking results, fixed line loss abnormal backtracking results and variable line loss abnormal backtracking results; According to the technical line loss abnormal backtracking result, the management line loss abnormal backtracking result, the fixed line loss abnormal backtracking result and the variable line loss abnormal backtracking result, the line loss abnormal backtracking map is generated.

4. The digital twin driven power distribution network line loss analysis method of claim 1, wherein, According to the multi-dimensional line loss risk prediction model, the distribution network adjustment guide space is established by analyzing the first distribution network control adjustment domain, including: According to the multi-dimensional line loss risk prediction model, technical line loss risk iterative optimization is performed on the first distribution network control adjustment domain to obtain a first distribution network adjustment guide vector; According to the multi-dimensional line loss risk prediction model, management line loss risk iterative optimization is performed on the first distribution network control adjustment domain to obtain a second distribution network adjustment guide vector; According to the multi-dimensional line loss risk prediction model, fixed line loss risk iterative optimization is performed on the first distribution network control adjustment domain to obtain a third distribution network adjustment guide vector; According to the multi-dimensional line loss risk prediction model, variable line loss risk iterative optimization is performed on the first distribution network control adjustment domain to obtain a fourth distribution network adjustment guide vector; The first distribution network adjustment guide vector, the second distribution network adjustment guide vector, the third distribution network adjustment guide vector and the fourth distribution network adjustment guide vector are added to the distribution network adjustment guide space.

5. The digital twin driven power distribution network line loss analysis method of claim 1, wherein, According to the multi-dimensional line loss risk prediction model, technical line loss risk iterative optimization is performed on the first distribution network control adjustment domain to obtain a first distribution network control adjustment domain, including: According to the first distribution network adjustment guide vector, difference analysis is performed on the first distribution network control adjustment domain to obtain a first difference characteristic particle swarm; According to the first difference characteristic particle swarm, multiple mutations are performed to obtain a first reproduction guide particle swarm, and the first distribution network control adjustment domain is guided to reproduce according to the first reproduction guide particle swarm to obtain a first distribution network adjustment reproduction domain; According to the second distribution network adjustment guide vector, the first distribution network control adjustment domain is guided to reproduce to obtain a second distribution network adjustment reproduction domain; According to the third distribution network adjustment guide vector, the first distribution network control adjustment domain is guided to reproduce to obtain a third distribution network adjustment reproduction domain; According to the fourth distribution network adjustment guide vector, the first distribution network control adjustment domain is guided to reproduce to obtain a fourth distribution network adjustment reproduction domain; According to the first distribution network adjustment reproduction domain, the second distribution network adjustment reproduction domain, the third distribution network adjustment reproduction domain and the fourth distribution network adjustment reproduction domain, the first distribution network control adjustment domain is expanded to generate the second distribution network control adjustment domain.

6. The digital twin driven power distribution network line loss analysis method of claim 1, wherein, According to the multi-dimensional line loss risk prediction model, multi-level optimization is performed on the second distribution network control adjustment domain to determine a distribution network control optimization result, including: According to the multi-dimensional line loss risk prediction model, line loss risk constraint optimization is performed on the second distribution network control adjustment domain to obtain a third distribution network control adjustment domain; According to the multi-dimensional line loss risk weight, a global line loss risk analysis model is constructed; According to the global line loss risk analysis model, global line loss risk analysis is performed on the third distribution network control adjustment domain to obtain a global line loss risk sequence; Based on the global line loss risk sequence, according to the global line loss risk threshold, the third distribution network control adjustment domain is optimized and identified to obtain a fourth distribution network control adjustment domain; According to the fourth distribution network control adjustment domain, energy consumption minimization optimization is performed to generate the distribution network control optimization result.

7. The digital twin driven power distribution network line loss analysis method of claim 6, wherein, According to the multi-dimensional line loss risk prediction model, line loss risk constraint optimization is performed on the second distribution network control adjustment domain to obtain a third distribution network control adjustment domain, including: According to the second distribution network control adjustment domain, a bth distribution network control adjustment scheme is extracted, b is a positive integer; obtain a bth line loss risk prediction matrix according to the multi-dimensional line loss risk prediction model based on the bth power distribution network control adjustment scheme; construct a line loss risk constraint matrix according to the technical line loss risk constraint, the management line loss risk constraint, the fixed line loss risk constraint and the variable line loss risk constraint; if the bth line loss risk prediction matrix meets the line loss risk constraint matrix, add the bth power distribution network control adjustment scheme to the third power distribution network control adjustment domain; continue to perform line loss risk constraint optimization on the second power distribution network control adjustment domain according to the multi-dimensional line loss risk prediction model based on the line loss risk constraint matrix, and update the third power distribution network control adjustment domain.

8. The digital twin driven power distribution network line loss analysis method of claim 1, wherein, The multi-dimensional line loss risk prediction model includes a multi-dimensional line loss risk prediction index, and the multi-dimensional line loss risk prediction index includes a technical line loss risk, a management line loss risk, a fixed line loss risk and a variable line loss risk.

9. The digital twin driven power distribution network line loss analysis method of claim 6, wherein, The multi-dimensional line loss risk weight includes a technical line loss risk weight, a management line loss risk weight, a fixed line loss risk weight and a variable line loss risk weight.

10. A digital twin driven power distribution network line loss analysis system characterized in that, The system is used to perform the digital twin driven power distribution network line loss analysis method according to any one of claims 1-9, and the system includes: a feature fitting module configured to perform multi-dimensional line loss feature fitting on the power distribution network based on a digital twin component according to a power distribution network control strategy, and establish a multi-dimensional line loss feature vector; an abnormality tracing module configured to perform abnormality tracing according to the multi-dimensional line loss feature vector, and establish a line loss abnormality tracing map; an adjustment module configured to adjust the power distribution network control strategy according to the line loss abnormality tracing map, and obtain a first power distribution network control adjustment domain; a guidance analysis module configured to perform power distribution network adjustment guidance analysis on the first power distribution network control adjustment domain according to a multi-dimensional line loss risk prediction model, and establish a power distribution network adjustment guidance space; a guidance propagation module configured to perform multi-level guidance propagation on the first power distribution network control adjustment domain according to the power distribution network adjustment guidance space, and establish a second power distribution network control adjustment domain; an optimization management module configured to perform multi-level optimization on the second power distribution network control adjustment domain according to the multi-dimensional line loss risk prediction model, determine a power distribution network control optimization result, and perform optimization management on the power distribution network based on the power distribution network control optimization result.

Images